Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!zaphod.mps.ohio-state.edu!pacific.mps.ohio-state.edu!linac,mp.cs.niu.edu!ux1.cso.uiuc.edu!rs6000.cmp.ilstu.edu!ejbehr
From: ejbehr@rs6000.cmp.ilstu.edu (Eric Behr)
Newsgroups: comp.lang.pascal
Subject: Storage efficiency
Message-ID: <1991Mar18.222538.5562@rs6000.cmp.ilstu.edu>
Date: 18 Mar 91 22:25:38 GMT
Reply-To: ejbehr@rs6000.cmp.ilstu.edu (Eric Behr)
Organization: Central Illinois Surfing Club
Lines: 21

I'm writing a program for rather exotic matrix computations (quick & dirty
research tool, nothing useful to an ordinary person). It will have to deal
with potentially large matrices (containing structured variables, two
longints each), and a lot of them at that. Since I want parts of it to do
data preparation which would run on a Mac, I want to make it efficient in
terms of storage. Here is the question: what do typical implementations do
when a large parameter is passed as a var? does this work exactly like
passing a pointer, or does the procedure somehow use more memory that way?

Similarly, is there a difference between passing the result (e.g. in a
procedure that multiplies two large matrices) as a var parameter, versus
having the function return it as a pointer to a matrix?

If it matters, I'm using Think Pascal on the Mac and IBM Pascal compiler
under AIX.

Sorry if this is trivial or confusing - I admit I haven't thought about it
too much. Thanks - E.
-- 
Eric Behr, Illinois State University, Mathematics Department
Internet: ejbehr@rs6000.cmp.ilstu.edu    Bitnet: ebehr@ilstu